Balanced Thinking: Improving Chain of Thought Training in Vision Language Models

• URL: http://arxiv.org/abs/2603.18656v1
• Date: Thu, 19 Mar 2026 09:21:49 GMT
• Title: Balanced Thinking: Improving Chain of Thought Training in Vision Language Models
• Abstract summary: Multimodal reasoning in vision-language models relies on a two-stage process: supervised fine-tuning and reinforcement learning.<n>We propose SCALe, which explicitly separates supervision over reasoning and answer segments using dynamic, length-independent weighting.<n>Results show that SCALe consistently improves accuracy over vanilla SFT and matches the performance of the full two-phase SFT + GRPO pipeline.
• Score: 6.47886000593456
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Multimodal reasoning in vision-language models (VLMs) typically relies on a two-stage process: supervised fine-tuning (SFT) and reinforcement learning (RL). In standard SFT, all tokens contribute equally to the loss, even though reasoning data are inherently token-imbalanced. Long <think> traces overshadow short but task-critical <answer> segments, leading to verbose reasoning and inaccurate answers. We propose SCALe (Scheduled Curriculum Adaptive Loss), which explicitly separates supervision over reasoning and answer segments using dynamic, length-independent weighting. Unlike vanilla SFT, which overweights the <think> segment, SCALe-SFT gradually shifts the focus from <think> to <answer> throughout training via a cosine scheduling policy, encouraging concise and well-grounded reasoning. We evaluate SCALe across diverse benchmarks and architectures. Results show that SCALe consistently improves accuracy over vanilla SFT and matches the performance of the full two-phase SFT + GRPO pipeline while requiring only about one-seventh of the training time, making it a lightweight yet effective alternative. When combined with GRPO, SCALe achieves the best overall performance, highlighting its value both as a standalone method and as a strong foundation for reinforcement refinement.

Related papers

• Is Next-Chunk Reasoning RL Really Better than SFT? Revisiting Training Strategies under no-CoT Data [48.92748013043833]
  We show that Mixed SFT achieves a clearly higher post-RLVR performance ceiling than next-chunk reasoning RL.<n>We also show that higher pre-RLVR accuracy does not necessarily translate into higher post-RLVR accuracy.
  arXiv  Detail & Related papers  (2026-08-24T13:47:45Z)
• Stop Thinking, Start Looking: Efficient Post-Training for Multimodal Document Question Answering via Reasoning-Free Alignment [2.8730465903425877]
  We introduce Perception-RFT, a training framework that applies Group Relative Policy Optimization to multimodal document QA.<n>We find that reasoning-enabled models suppress their reasoning traces during training, converging to direct perception-based policies at the 4B parameter scale.<n>We show that an early SFT$rightarrow$RL transition achieves comparable precision with 65% less training data.
  arXiv  Detail & Related papers  (2026-07-16T07:43:39Z)
• Debunk the Myth of SFT Generalization [13.700645417996412]
  A prevailing view holds that supervised fine-tuning (SFT) fails to generalize, whereas reinforcement learning (RL) attains broader robustness.<n>We show that much of SFT's perceived failure stems from frozen-prompt artifacts.<n>We ask whether SFT can generalize strictly harder tasks.
  arXiv  Detail & Related papers  (2025-09-30T20:01:09Z)
• The Synergy Dilemma of Long-CoT SFT and RL: Investigating Post-Training Techniques for Reasoning VLMs [66.17068546293487]
  Large vision-language models (VLMs) increasingly adopt post-training techniques such as long chain-of-thought (CoT) supervised fine-tuning (SFT) and reinforcement learning (RL) to elicit sophisticated reasoning.<n>We present a systematic investigation into the distinct roles and interplay of long-CoT SFT and RL across multiple multimodal reasoning benchmarks.<n>We find that SFT improves performance on difficult questions by in-depth, structured reasoning, but introduces verbosity and degrades performance on simpler ones.
  arXiv  Detail & Related papers  (2025-07-10T09:05:49Z)
• RL for Reasoning by Adaptively Revealing Rationales [36.50924054394857]
  Supervised fine-tuning (SFT) relies on dense ground-truth labels, which become increasingly costly as sequence length grows.<n>We address this by adaptive backtracking (AdaBack), a per-sample curriculum learning algorithm that reveals only a partial prefix of the target output during training.<n>We show that our adaptive curriculum over partial answers reliably solves problems that are otherwise intractable.
  arXiv  Detail & Related papers  (2025-06-22T17:46:14Z)
• Skip-Thinking: Chunk-wise Chain-of-Thought Distillation Enable Smaller Language Models to Reason Better and Faster [51.89995713333108]
  Chain-of-thought (CoT) distillation allows a large language model (LLM) to guide a small language model (SLM) in reasoning tasks.<n>Existing methods train the SLM to learn the long rationale in one iteration.<n>We propose chunk-wise training (CWT), which uses a search to divide the rationale into internal semantically coherent chunks.
  arXiv  Detail & Related papers  (2025-05-24T11:04:52Z)
• Compile Scene Graphs with Reinforcement Learning [69.36723767339001]
  Next-token prediction is the fundamental principle for training large language models (LLMs)<n>We introduce R1-SGG, a multimodal LLM (M-LLM) initially trained via supervised fine-tuning (SFT) on the scene graph dataset.<n>We design a set of graph-centric rewards, including three recall-based variants -- Hard Recall, Hard Recall+Relax, and Soft Recall.
  arXiv  Detail & Related papers  (2025-04-18T10:46:22Z)
• Exploring the Effect of Reinforcement Learning on Video Understanding: Insights from SEED-Bench-R1 [53.894789613838654]
  We introduce SEED-Bench-R1, a benchmark designed to evaluate post-training methods for MLLMs in video understanding.<n>It includes intricate real-world videos and complex everyday planning tasks in the format of multiple-choice questions.<n>Using Qwen2-VL-Instruct-7B as a base model, we compare RL with supervised fine-tuning (SFT)<n>Our detailed analysis reveals that RL enhances visual perception but often produces less coherent reasoning chains.
  arXiv  Detail & Related papers  (2025-03-31T17:55:23Z)
• Sketch-of-Thought: Efficient LLM Reasoning with Adaptive Cognitive-Inspired Sketching [64.74765550805024]
  Chain-of-Thought prompting elicits step-by-step problem solving, but often at the cost of excessive verbosity in intermediate outputs.<n>We propose Sketch-of-Thought (SoT), a prompting framework that integrates cognitively inspired reasoning paradigms with linguistic constraints.<n>SoT achieves token reductions of up to 84% with minimal accuracy loss across 18 reasoning datasets.
  arXiv  Detail & Related papers  (2025-03-07T06:57:17Z)
• Intuitive Fine-Tuning: Towards Simplifying Alignment into a Single Process [19.986235452236272]
  Supervised Fine-Tuning (SFT) and Preference Optimization (PO) are key processes for aligning Language Models (LMs) with human preferences post pre-training.<n>We introduce Intuitive Fine-Tuning (IFT) to integrate SFT and PO into a single process.<n>IFT performs comparably or even superiorly to SFT and some typical PO methods across several tasks.
  arXiv  Detail & Related papers  (2024-05-20T08:23:28Z)
• ReFT: Reasoning with Reinforced Fine-Tuning [9.80361828538909]
  We propose a simple yet effective approach called Reinforced Fine-Tuning (ReFT) to enhance the generalizability of learning LLMs for reasoning.<n>ReFT first warmups the model with SFT, and then employs on-line reinforcement learning, specifically the PPO algorithm in this paper.<n>Experiments on GSM8K, MathQA, and SVAMP datasets show that ReFT significantly outperforms SFT.
  arXiv  Detail & Related papers  (2024-01-17T04:43:21Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.